Task-specific relationships between error-related ERPs and behavior: Flanker, Stroop, and Go/Nogo tasks.

Area of Science:

• Cognitive Neuroscience
• Psychology
• Neuroscience

Background:

• Performance monitoring is crucial for adaptive behavior, often studied via event-related brain potentials (ERPs) and behavioral metrics like response time (RT) and accuracy.
• Previous research shows inconsistent links between error-related ERPs and subsequent behavior, potentially due to task variations.
• Understanding how different tasks influence these ERP-behavior relationships is essential for robust interpretation of performance monitoring findings.

Purpose of the Study:

• To investigate whether the cognitive task moderates the relationship between error-related ERPs and behavioral outcomes (RT, accuracy).
• To determine if these ERP-behavior relationships are consistently observed when tasks and stimuli are treated as random effects.

Main Methods:

• Recorded ERPs and behavioral data (RTs, accuracy) from 180 participants across three distinct tasks: flanker, Stroop, and Go/Nogo.
• Analyzed the relationship between previous-trial error-related negativity (ERN) and error positivity (Pe) with subsequent trial performance.
• Examined the robustness of these relationships by treating tasks and stimuli as random effects.

Main Results:

• A significant task-specific effect was found: larger ERN predicted longer RTs and higher accuracy in flanker and Stroop tasks, but not Go/Nogo.
• Larger Pe predicted faster RTs and reduced RT variability in Stroop and Go/Nogo tasks, but not in the flanker task.
• ERP-behavior relationships were not observed when tasks and stimuli were treated as random effects, highlighting variability.

Conclusions:

• The observed ERP-behavior relationships are task-dependent, indicating that task selection critically influences performance monitoring measures.
• Findings underscore the necessity of considering the specific task employed when interpreting and comparing results across studies on performance monitoring.
• The lack of robust relationships under random effects suggests that task context is a primary moderator, not a secondary nuisance variable.